Storing Data On A Single Atom

In the electronics industry, the march of time brings with it a reduction in size. Our electronic devices, while getting faster, better and cheaper, also tend to get smaller. One of the main reasons for this is the storage medium for binary data gets smaller and more efficient. Many can recall the EPROM, which is about the size of your thumb. Today we walk around with SD cards that can hold an order of magnitude more data, which can fit on your thumb’s nail.

Naturally, we must ask ourselves where the limit lies. Just how small can memory storage get? How about a single atom! IBM along with a handful international scientists have managed to store two bits of information on two pairs of holmium atoms. Using a scanning tunneling microscope, they were able to write data to the atoms, which held the data for an extended period of time.

Holmium is a large atom, weighing in at a whopping 67 AMU. It’s a rare earth metal from the lanthanide series on the periodic table. Its electron configuration is such that many of the orbiting electrons are not paired. Recall from our article on the periodic table that paired electrons must have opposite spin, which has the unfortunate consequence of causing the individual magnetic fields to cancel. The fact that holmium has so many unpaired electrons makes it ideal for manipulation.

While you won’t be seeing atom-level memory on the next Raspberry Pi, it’s still neat to see what the future holds.

Thanks to [Itay] for the tip!

Via Gizmodo.

Google Machine Learning Made Simple(r)

If you’ve looked at machine learning, you may have noticed that a lot of the examples are interesting but hard to follow. That’s why [Jostmey] created Naked Tensor, a bare-minimum example of using TensorFlow. The example is simple, just doing some straight line fits on some data points. One example shows how it is done in series, one in parallel, and another for an 8-million point dataset. All the code is in Python.

If you haven’t run into it yet, TensorFlow is an open source library from Google. To quote from its website:

TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) communicated between them. The flexible architecture allows you to deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device with a single API. TensorFlow was originally developed by researchers and engineers working on the Google Brain Team within Google’s Machine Intelligence research organization for the purposes of conducting machine learning and deep neural networks research, but the system is general enough to be applicable in a wide variety of other domains as well.

Continue reading “Google Machine Learning Made Simple(r)”

Hush Those Old-Fashioned Phones

Most people hate unsolicited calls, and it’s worse in the dead of night when we’re all trying to sleep. Smartphones are easy to configure to block nuisance calls, but what if you need a solution for your Plain Old Telephone System (POTS)? [Molecular Descriptor] has built a system to invisibly stop landline phones ringing after hours.

The basic principle relies on an analog circuit that detects the AC ringing signal from the phone network, and then switches in an impedance to make the phone company think the phone has been picked up. The circuit is able to operate solely on the voltage from the phone line itself, thanks to the use of the LM2936 – a regulator with an ultra-low quiescent current. It’s important if you’re going to place a load on the phone line that it be as miniscule as possible, otherwise you’ll have phone company technicians snooping around your house in short order wondering what’s going on.

The aforementioned circuitry is just to block the phone line. To enable the system to only work at night, more sophistication was needed. An Arduino Mega was used to program an advanced RTC with two alarm outputs, and then disconnected. The RTC is then connected to a flip-flop which connects the blocking circuit only during the requisite “quiet” hours programmed by the Arduino. The RTC / flip-flop combination is an elegant way of allowing the circuit to remain solely powered by the phone line in use, as they use far less power when properly configured than a full-blown microcontroller.

It’s a cool project, with perhaps the only pitfall being that telecommunications companies aren’t always cool with hackers attaching their latest homebrewed creations to the network. Your mileage may vary. For more old-school telephony goodness, check out this home PBX rig.

Arpeggio – The Piano SuperDroid

I never had the musical talent in me. Every now and then I would try to pick up a guitar or try and learn the piano, romanticising a glamorous career out of it at some point. Arpeggio – the Piano SuperDroid (YouTube, embedded below) sure makes me glad I chose a different career path. This remarkable machine is the brain child of [Nick Morris], who spent two years building it.

Although there are no detailed technical descriptions yet, at its heart this handsome robot consists of a set of machined ‘fingers’ connected to a set of actuators — most likely solenoids . The solenoids are controlled by proprietary software that combines traditional musical data with additional parameters to accurately mimic performances by your favourite pianists, right in your living room. Professional pianists, who were otherwise assuming excellent job security under Skynet, clearly have to reconsider now.

Continue reading “Arpeggio – The Piano SuperDroid”

Laser Cutting A 3D Printer

The concept of self-replicating 3D printers is a really powerful one. But in practice, there are issues with the availability and quality of the 3D-printed parts. [Noyan] is taking a different approach by boostrapping a 3D printer with laser-cut parts. There are zero 3D-printed parts in this project. [Noyan] is using acrylic for the frame and the connecting mechanisms that go into the machine.

The printer design chosen for the project is the Prusa i3. We have certainly seen custom builds of this popular design before using laser-cut plywood for the frame. Still, these builds use 3D-printed parts for some of the more complicated parts like the extruder carriage and motor brackets. To the right is the X-carriage mechanism. It is complicated but requires no more than 6 mm and 3 mm acrylic stock and the type of hardware traditionally associated with printer builds.

With the proof of concept done, a few upgrades were designed and printed to take the place of the X-axis parts and the belt tensioner. But hey, who doesn’t get their hands on a 3D printer and immediately look for printable solutions for better performance?

We first saw a laser-cut RepRap almost nine years ago! That kit was going to run you an estimated $380. [Noyan] prices this one out at under $200 (if you know someone with a laser cutter), and of course you can get a consumer 3D printer at that price point now. Time has been good to this tool.

The Inventions Of Arthur Paul Pedrick

We hear a lot about patent portfolios when we scan our morning dose of tech news stories. Rarely a day passes without news of yet another legal clash between shady lawyers or Silicon Valley behemoths, either settling spats between multinationals or the questionable activities of patent trolls.

These huge and well-heeled organisations hold many patents, which they gather either through their staff putting in the hard work to make the inventions, or by acquisition of patents from other inventors. It is not often that a large quantity of patents are amassed by any other means, for example by an individual.

There is one prolific individual inventor and holder of many patents though. He achieved notoriety not through his inventions being successful, but through their seeming impracticability while conforming to the rules of the patent system. His name was [Arthur Paul Pedrick], and he was a retired British patent examiner who filed a vast number of eccentric patents from the early 1960s until his death in the mid 1970s, all of which stretched the boundaries of practicality.

His subject matter was varied, but included a significant number of transport inventions as well as innovations in the field of energy and nuclear physics. We wish there was room to feature them all on these pages, but sadly they are so numerous that it is difficult even to pick the selection we can show you. So sit down, and enjoy the weird and wonderful world of [Pedrick] innovations.

Continue reading “The Inventions Of Arthur Paul Pedrick”

[Ashhar Farhan]’s Done It Again!

If you are a regular follower of these pages as well as a radio amateur, you may well have heard of [Ashhar Farhan, VU2ESE]. He is the designer of the BitX, a simple single-sideband transceiver that could be built for a very small outlay taking many of its components from a well-stocked junk box.

In the years since the BitX’s debut there have been many enhancements and refinements to the original, and it has become something of a standard. But it’s always been a single-band rig, never competing with expensive commercial boxes that cover the whole of the available allocations.

With his latest design, he’s changed all that. The uBITX (Micro-BITX when spoken aloud), is an SSB and CW transceiver that covers all of the HF amateur bands, and like the original is designed for the home constructor on a budget. It shows its heritage in the use of bi-directional amplifiers, but diverges from the original with a 45 MHz first IF and an Arduino/SI5351 clock generator in the place of a VFO. It looks to be an excellent design in the spirit of the original, and we can’t wait to see them in the wild.

He’s put up a YouTube video which we’ve placed below the break. His write-up is extensive and fascinating, but it is his closing remarks which sum up the project and the reason why you should build one. We don’t often reproduce entire blocks of text, but this one says it so well:

As a fresh radio amateur in the 80s, one looked at the complex multiband radios of the day with awe. I remember seeing the Atlas 210x, the Icom 720 and Signal One radios in various friends’ shacks. It was entirely out of one’s realm to imagine building such a general coverage transceiver in the home lab.

Devices are now available readily across the globe through online stores, manufacturers are more forthcoming with their data. Most importantly, online communities like the EMRFD’s Yahoo group, the BITX20’s groups.io community etc have placed the tribal knowledge within the grasp of far flung builders like I am.

One knows that it was just a matter of breaking down everything into amplifiers, filters, mixers and oscillators, but that is just theory. The practice of bringing a radio to life is a perpetual ambition. The first signal that the sputters through ether, past your mess of wires into your ears and the first signal that leaps out into the space from your hand is stuff of subliminal beauty that is the rare preserve of the homebrewer alone.

At a recent eyeball meet, our friend [Dev(VU2DEV)] the famous homebrewer said “Now is the best time to be a homebrewer”. I couldn’t disagree.

If you build a uBITX, please share it with us!

Continue reading “[Ashhar Farhan]’s Done It Again!”